Related Experiment Video
Updated: Aug 5, 2026
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
Single-cell immune subtyping of basophils in head and neck squamous cell carcinoma and a prognostic model based on
Xinyu Zhao1,2, Hongfa Wang1,2, Huang Huang1,2
1Center for Rehabilitation Medicine, Department of Anesthesiology, Zhejiang Provincial People's Hospital (Affiliated People's Hospital, Hangzhou Medical College), Hangzhou, Zhejiang, 310014, China.
Background:
The precise contribution of basophils to cancer development and advancement remains undefined. Our research sought to delineate the immune subtypes of basophils in head and neck squamous cell carcinoma (HNSCC) and to develop a robust prognostic model.
Methods:
Single-cell and bulk transcriptomic data from the TCGA-HNSC and the GEO (GSE41613, GSE139324) were integrated. Basophil subtypes and their signature genes were identified. Samples were subjected to unsupervised consensus clustering based on basophil-related differentially expressed genes to define immune subtypes. A prognostic model was constructed and its performance validated in external datasets. The biological function of the key gene DHRS2 was investigated through in vitro experiments.
Results:
Single-cell analysis identified heterogeneous basophil populations in HNSCC, classified into distinct subtypes. Cluster 1 displayed a low immune infiltration phenotype. Using basophil signature genes, TCGA-HNSC samples were stratified into two subtypes with significant survival differences. A 12-gene prognostic model with good predictive performance was developed. High-risk patients showed high tumor purity along with downregulated immune checkpoint and HLA gene expression, indicative of active immune evasion. Experimentally, the key model gene DHRS2 was highly expressed in HNSCC, and its knockdown suppressed cell proliferation, migration, and invasion while inducing apoptosis.
Conclusion:
The study characterizes basophil immune subtypes in HNSCC and establishes a robust basophil-related prognostic model that effectively identifies high-risk patients, suggesting the potential pro-tumorigenic gene DHRS2 as a therapeutic target in HNSCC.
